Prevalence and Clinical Correlation of Anti-Phospholipid–Binding Protein Antibodies in Anticardiolipin-Negative Patients With Systemic Lupus Erythematosus and Women With Unexplained Recurrent Miscarriages

Abstract

Anti-phospholipid antibodies (aPL) are a heterogeneous group of autoantibodies, the presence of which is associated with thrombotic events and miscarriage. To establish whether antibodies directed against phospholipid-binding plasma proteins such as beta(2)-glycoprotein I (beta(2)GPI), prothrombin (PT), and annexin V (Anx V) constitute a risk factor for thromboembolism in patients with systemic lupus erythematosus (SLE) and for miscarriage in women with recurrent pregnancy loss (RPL), independently of the presence of the classic anticardiolipin (aCL) antibodies, and whether their determination together with that of aCL would help to increase the diagnostic sensitivity of aPL tests. The prevalence of various antibodies directed toward phospholipids (CL and other anionic phospholipids [APL]) and phospholipid-binding proteins (beta(2)GPI, PT, and Anx V) was determined by immunoenzymatic methods in 311 serum samples. Twenty-five patients with aCL-positive primary anti-phospholipid syndrome (pAPS); 89 patients with SLE, 23 of whom had thrombotic complications (SLE/APS) and 66 of whom had no thrombosis; and 77 women with unexplained recurrent pregnancy loss comprised our study group. One hundred twenty healthy subjects matched for age and sex were studied as the control group. Immunoglobulin (Ig) G and/or IgM aAPL, anti-beta(2)GPI, anti-PT, and IgG anti-Anx V antibodies were detected in 25 (100%), 20 (80%), 15 (60%), and 6 (24%), respectively, of the 25 aCL-positive pAPS patients; IgG and/or IgM aCL, aAPL, anti-beta(2)GPI, anti-PT, and IgG anti-Anx V antibodies were detected in 33 (37%), 42 (47%), 31 (35%), 40 (45%), and 12 (13%) of the 89 SLE patients, respectively. Of the 56 SLE patients who proved to be aCL negative, anti-beta(2)GPI was present in 3 patients (5%), anti-PT in 13 (23%) patients, and anti-Anx V in 5 (9%) patients. In the subset of 23 SLE/APS patients, IgG anti-PT prevalence was higher than that of the other autoantibodies (87% vs 70% aCL, 66% aAPL, 57% anti-beta(2)GPI, and 4% anti-Anx V), and in 26% of cases, IgG anti-PT was the only antibody present. Anti-PT had a slightly lower specificity than aCL (46% vs 49%); however, the occurrence of both antibodies brought the specificity to 92.4%. The highest risk for thrombosis in SLE patients was associated with the presence of IgG anti-PT antibody (odds ratio [OR] 15.3, P < .001, vs 6.5 aCL, 3.5 aAPL, 3.4 anti-beta(2)GPI, 0.2 anti-Anx V). Fifty-one of the 77 women with recurrent pregnancy loss were negative for all antibodies investigated; the prevalence of IgG and/or IgM aCL, aAPL, anti-beta(2)GPI, anti-PT, and IgG anti-Anx V antibodies was 6% (5), 12% (9), 6% (5), 16% (12), and 17% (13), respectively. Of the 67 aCL-negative women, none had anti-beta(2)GPI antibodies, 7 (11%) were anti-PT positive, and 13 (19%) were anti-Anx V positive. In the subgroup of 26 recurrent pregnancy loss patients who had at least one antibody, anti-Anx V was present in 50% of cases (in 42% as the sole antibody) and was the only antibody significantly associated with miscarriage (P = .02). The results of this study indicate that it is useful to measure anti-PT antibodies in addition to the more widely used aCL and anti-beta(2)GPI antibodies in the prognostic evaluation of SLE patients for the risk of thrombosis, and the results also confirm that anti-Anx V antibodies may play an important role in recurrent pregnancy loss.